Security audit

图表工具免费版

Security checks for vulnerabilities and agentic risk

Overview

The skill is mainly a Mermaid diagram helper, but it asks for read and command access while advertising broader data, file, API, network, and retrieval behavior than its purpose needs.

Review before installing. Use this only for Mermaid diagram creation or clearly requested PNG export, and avoid letting it run unrelated data analysis, API calls, broad file processing, or network diagnostics unless you explicitly intend that.

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration

Findings (5)

Intent-Code Divergence

Medium

Confidence: 94% confidence
Finding: The skill claims no extra API key or external networking is required, yet later discusses API-key handling, HTTPS, SSL validation, ping-based connectivity tests, and proxy/firewall troubleshooting. This inconsistency can mislead an agent or user about the skill’s true trust boundary and may normalize unnecessary network activity in a skill whose stated purpose is local diagram generation.

Description-Behavior Mismatch

Medium

Confidence: 97% confidence
Finding: The manifest presents a Mermaid diagram generator, but the documented core functions expand into generic file handling, API integration, command execution, and information retrieval. This capability inflation widens the apparent authority of the skill and can cause agents to invoke it for unrelated tasks, increasing the chance of unnecessary file access or command execution.

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: Advertising external API integration and broad information retrieval in a diagram-generation skill creates unjustified functional scope beyond the user’s likely expectation. In a skill with exec permission, such vague expansion is dangerous because it legitimizes additional data access or outbound interactions that are not necessary for producing Mermaid diagrams.

Vague Triggers

High

Confidence: 93% confidence
Finding: The top-level description says to use the skill for data analysis, report generation, statistical insights, and data visualization, which is far broader than a Mermaid diagram helper. Overbroad triggers increase the likelihood that an agent will auto-select this skill in unrelated contexts, exposing exec/read capabilities where they are unnecessary.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The documented trigger condition is vague and does not clearly separate supported diagram use cases from unsupported analytics-style requests. Ambiguous activation boundaries can cause unintended invocation, especially because the skill has read and exec permissions and includes instructions that imply broader operational behavior.

Static analysis

No suspicious patterns detected.